Here's the ugly timeline nobody puts on the dashboard. A lead hits "Submit" at 2:04:12. Your CRM shows the contact created at 2:04:14. Your first text lands on their phone at 2:19:41. Fifteen and a half minutes evaporated between a warm intent signal and the first human-shaped touch, and if you asked five people on your team where those minutes went, you'd get five shrugs.
Full disclosure: I work for Ready, and we sell both SMS and a power dialer with speed-to-lead auto-dial, so I have a stake in you caring about this. But the audit below is provider-agnostic. Most of these delays are in your own plumbing, and you can find and kill them this afternoon regardless of who you send with.
The general finding — well-worn by now — is that reply and connect rates fall off a cliff as the minutes pile up. We covered the shape of that curve in The 60-Second Window. This post is about where the 15 minutes physically hide.
The hops, laid out
Every form-to-first-touch path is a chain of handoffs. Each one adds latency. Here's the typical chain and roughly what each hop costs when it's misconfigured:
| Hop | What happens | Typical delay when broken |
|---|---|---|
| 1. Form submit → contact create | Form provider posts to your CRM | 0–3 sec (usually fine) |
| 2. Contact create → workflow trigger | CRM decides a workflow should fire | 0 sec native / 1–3 min via middleware |
| 3. Workflow wait step | A "Wait" action someone left in | 1–15 min, silent |
| 4. Business-hours / quiet-hours gate | Send held until "office hours" | 0 min to 14 hours |
| 5. Provider queue → carrier | Message queued, throughput-limited | 0–90 sec native / 2 min+ via Zapier |
| 6. Carrier → handset | Out of your hands | 1–10 sec |
The two you control least — hops 1 and 6 — are almost never the problem. The middle four are where your 15 minutes live. Let's walk each.
Hop 2: the middleware tax
If your form fires a webhook into Zapier, Make, or a custom Lambda before anything reaches your messaging layer, you've inherited a queue. Zapier's standard polling and task queues can add a minute or two before your "instant" automation even wakes up. On a slow tier it's worse.
I wrote up the measured version of this in Routing SMS Through Zapier Adds a 2-Minute Delay. The short version: every intermediary between the form and the send is a place where a message sits in someone else's queue waiting for a worker.
Fix: collapse the chain. If you're in GoHighLevel, Ready's native OAuth sync means inbound and outbound messages move through the platform directly rather than being relayed by a third party — the workflow that fires the text and the messaging layer are the same system, so there's no relay hop to queue behind. If you're not in GHL, the principle holds: send directly from the tool that already has the contact, not through a bolt-on.
Hop 3: the "Wait" step nobody remembers adding
This is the single most common culprit and the easiest to miss, because a workflow with a 10-minute Wait step looks completely normal in the builder. Someone added it during testing to avoid spamming themselves, or copied a template that had it baked in, or throttled the branch during a deliverability scare and never took it back out.
Go open your new-lead workflow right now. Read every step between the trigger and the SMS action. If there's a Wait, Drip, or Delay action before the first text, that's your leak. A 10-minute wait plus a 2-minute middleware queue and you're already at 12 minutes before the carrier ever sees the message.
Fix: the first outbound touch should be the very first action after the trigger — no wait, no condition that can stall, no "if email opened" branch it has to evaluate first. Put the delays after the first text, in the follow-up cadence, where they belong. If you need branch logic, split it so the immediate send happens unconditionally and the branching happens downstream. Related failure mode: sending the SMS and email on the same step, which we cover in Why That Doubles Your Unsub Rate.
Hop 4: the business-hours gate that swallows afternoons
This one is sneaky because it's supposed to be there. Quiet-hours enforcement is a real compliance control — you don't want a lead form filled at 11:40 PM generating a text at 11:41 PM, because that's TCPA exposure. Ready holds sends outside permitted local hours automatically based on the recipient's area, and that's the correct default.
The failure is when the business-hours window is set too narrow and it's fighting your speed-to-lead, not protecting it. If your workflow says "only send between 9 and 5" and a lead comes in at 5:14 PM, that message gets parked until 9 AM tomorrow. Your 15-minute problem just became a 16-hour problem, and by morning the lead has filled out three competitors' forms and forgotten yours.
Fix: separate two different windows in your head.
- Legal quiet hours (roughly 8 AM–9 PM local, TCPA-driven) — keep these on, always.
- Your office hours — do not gate the first automated text to these. A same-second text at 5:14 PM is fine and legal; a same-second dial from a rep who's clocked out isn't going to happen anyway.
For everything that lands after your reps leave, you want a distinct after-hours path — an immediate text acknowledging the form, then a dial queued for when someone's back. We laid that cadence out in 62% of Leads Arrive After Your Reps Clock Out.
Hop 5: the provider queue and throughput ceiling
Even with a clean workflow, your send can wait in your provider's outbound queue. This matters most in bursts — a paid campaign spiking 300 form fills in ten minutes will throttle against your registered throughput limit, and the 250th lead's text goes out well after the 1st's.
Two things govern this: your 10DLC registration (higher trust score = higher daily throughput) and how directly the message reaches the carrier. Standard 10DLC — roughly $10/mo per brand and $20/mo per campaign — covers most senders. If you're consistently bursting large volumes, optional Brand Vetting ($40 Standard / $100 Enhanced, one-time) raises your throughput ceiling. Don't buy it unless your volume actually justifies it; most senders don't need it. The mechanics of registration are in What Is 10DLC.
Fix: register properly, and send directly rather than through a relay that adds its own queue on top of the carrier's.
The dial, not just the text
Everything above gets your text out fast. But the highest-converting speed-to-lead motion is a text and a dial in the first few minutes — one touch to land in the inbox, one to actually talk. We made that case in One Dial and One Text in the First 5 Minutes Beats Six Calls Over Three Days.
The latency audit for the dial is the same walk — trigger, wait step, hours gate, queue — but the payoff is bigger because a live conversation closes at a rate no text does. Ready's Power Dialer has speed-to-lead auto-dial on the Team plan ($69/agent/mo, $0.0375/min), which fires a call to a rep the moment a new lead lands, skipping the manual "notice the lead, open the record, dial" gap that quietly adds its own three minutes. Text-first-or-call-first is a real decision, not a default — we broke down which order wins by lead type.
Run the audit in 20 minutes
You don't need tooling for this. You need a stopwatch and one test lead.
- Submit your own form and write down the exact second.
- Watch the CRM — note when the contact appears. (Hop 1. Should be seconds.)
- Open the workflow and read it top to bottom. Flag any Wait/Delay/Drip before the first send. (Hop 3.)
- Check the hours gate on that send. Is it clamped to office hours instead of legal quiet hours? (Hop 4.)
- Note when the text actually lands on your phone. (Hops 5–6.)
- Subtract. The gap between contact-created and text-landed, minus a few seconds of carrier time, is entirely your own plumbing.
Do this at 2 PM and again at 5:15 PM. If the 5:15 test parks the message until tomorrow, you found your biggest leak and it isn't even latency — it's a gate closing the door.
The practical takeaway
Fifteen minutes almost never disappears into one dramatic failure. It's a minute of middleware, ten minutes of a forgotten Wait step, and an hours gate that shipped the whole thing to tomorrow. Each one is boring, silent, and fixable in a single sitting.
The move: collapse your relay hops, make the first text the first action after the trigger, gate only on legal quiet hours, and pair the text with an auto-dial so a rep is talking before the lead cools. If you want to see what direct-from-platform sending and speed-to-lead dialing look like together, the details are on the Ready product page, and you can start with 2,500 free credits — no card — at signup. Or just run the 20-minute audit above first. You'll probably find the fix before you find a new provider.